Timed neural Petri nets
N. Chamas, Lisa Anneberg, Ece Yaprak · 2002
This paper proposes the concept of timed neural Petri nets, which are isomorphic to neural net architecture. The timed neural Petri net concept combines the new orientation of the neural net with the modeling capabilities of the Petri Net. A big advantage to the utilization of Petri nets for modeling (especially in the neural case) is the proper representation of concurrent or parallel events. The neural net architecture techniques are applied to the Petri net, and the neural Petri net concept extends the basic Petri net definition. The paper is divided into four parts: introduction to neuron concept, timed neural Petri net (TNPN) development, analysis and example of the TNPN, and conclusions.>